Abstract: The road detection technique and navigation of vehicle is challenging in different road scenarios in the real urban environment, particularly only when perceptions sensors and lasers were used without position information. Several methods have been used to derive this concept in real time application where illumination of light is invariant. These methods present a novel real-time optimal drivable region and this methods are robust to shadows and lighting variations. Their system uses a multisensory scheme to cover the most drivable area in front of the vehicle. A colour based tracking system is used to get the optimal region on an unmarked road. This methods hold true in different environments and weather conditions. Using a stereophonic vision system of 3D parameters they propose a fast road detections from colour images. It is simple and can be applied on real time scenario. They propose system successfully which handles both structured and unstructured roads. They get better understanding of this concept to produce a result of better understanding following numeroustests.